ABSTRACT
This chapter introduces a robust image inpainting algorithm to determine a new filling order based on the structure priority value. Image inpainting is one such authoring technique for repairing or adjusting on-hand images. Image inpainting focuses on automatically repairing the missed regions of an image in a visually plausible way. The most important issue for image inpainting of large missing area is to determine the structure information of these areas, and repair them seamlessly from existing surrounding information. The inpainting algorithm concerns two major issues: appropriate filling order and adequate searching range. The proposed inpainting algorithm defines the new parameter, structure priority value, which can efficiently keep the structure information of an image. Most related studies use an exemplar-based method to keep the texture information of an image. The algorithm is only suitable for some structure distribution, and produces a blur effect in the repaired image.
